An experimental investigation on utilization of ladle refined furnace (LRF) slag in stabilizing clayey soil

Abstract

The contemporaneous world of development and urbanization generates extreme stresses on building infrastructures, and the use of soil as a load-bearing material is a prerequisite for such activity. However, if the foundation soil lacks enough bearing capacity and shear strength and shows excessive settlements, unescapable failures will occur in constructed infrastructures. As such, improvement or stabilization of soft clay soil is crucial before initiating any construction project to limit the potentiality of post-construction faults and damages. In this research, the efficiency of ladle refined furnace (LRF) slag has been investigated to stabilize the geotechnical properties of poor conditioned-soft clayey soil, which will also lessen the environmental degradation due to the unplanned dumping of LRF slag. Various experimental programs such as the Atterberg limit, consolidation, compaction, unconfined compressive strength, and California Bearing Ratio tests were carried out in this study. Experimental results revealed that adding LRF slag in different percentages augmented the strength and dry density of the original soil to a considerable extent, and the optimum percentage of LRF slag was 20%. 20% LRF content in the soil increased the UCS value of the soil by 219%, CBR value by 47%, and reduced consolidation settlement by 30%. Additionally, the test results were further evaluated by the SEM-EDS test.

1. Introduction

The properties of base soil often govern the strength and durability of a building or structure. Soft clay soil possessing poor properties may generate extensive distress on structural strength, which cannot be disregarded. These soils are expansive or dispersive in nature, having surprisingly delicate engineering properties, which makes the researchers worried [[1], [2], [3], [4], [5]]. The predictable damage cost caused by soil expansion surpasses the cost of damages from natural calamities, and numerous techniques are being espoused to lessen the penalties of such weak soil conditions [6]. Such properties of soil show their tendency to be changed in volume with the changes in moisture content [7,8], lower shear strengths, higher compressibility [2,[9], [10], [11], [12]], high susceptibility to shrinking and swelling [2,5,[13], [14], [15]], higher plasticity, and low permeability [12], which trigger the failure of bearing capacity and unwanted settlement. Exchanging or replacing the soft soil with superior property belonging soil is not always feasible, while stabilization has economic and environmental benefits for reinforcing such soil [16,17]. Besides various conventional methods, with the limited resources available, soft ground improvement inspires researchers to think about the circular economy for reusing industrial byproducts [[18], [19], [20]].

Moreover, the present urban dynamics motivate industries for newer types of products that generate waste byproducts that need to be managed sustainably. Researchers are trying their best to invent various techniques of soil improvement by reutilizing industrial waste byproducts to ensure not only the increase in strength of soil but also lessen the landfilling induced disquiets of those byproducts such as environmental concerns and costs, including treatment, transport, and land acquisition. Using different types of slag for soil stabilization or improvement is a well-known practice in geotechnical engineering to solve the problems of poor clay soil conditions and to manage the waste byproducts by reusing them. Ladle refined furnace (LRF) slag is an industrial byproduct from the steel-making industry, produced after the secondary refining of steel. Since LRF slag contains cementitious characteristics [2,[21], [22], [23]], it can be used as a soil stabilization material. In this research, the stabilizing effect of LRF slag has been utilized to improve the geotechnical parameters of soft clayey soil, reducing the environmental degradation due to the disposal of LRF slag.

1.1. Ladle refined furnace (LRF) slag and its uses

Ladle refined furnace (LRF) slag is a residue of the secondary metallurgy process [24] of steel manufacturing, which is also called ladle metallurgy furnace slag (LMS) [2]. The generation of LRF slag mainly depends on how the steel is refined, and hence the amount of LRF slag generation may vary with the degree of refinement in different types of steel plants. Maghool et al. [25] and Skaf et al. [26] stated that one ton of steel generates 30–50 kg and 60–80 kg of LRF slag, respectively, while Mahoutian & Shao [27] estimated that yearly about 30 million tons of LRF slag are produced in the world. According to Ahmad & Rahman [28], Bangladesh consumes about four million tons of steel yearly, produced and supplied by around 400 steel mills. BSRM, a well-reputed steel manufacturing company here, produces 12–15 kg LRF slag per ton of steel. The common destiny of LRF is the landfill [29,30], which is a constriction because of the inadequacy of lands and environmental issues. Other than landfilling LRF slag, researchers have been trying to point out different applications for it towards sustainability as it has environmental issues [31]. Due to the physicochemical properties of LRF slag, it has vast usage in civil engineering applications. LRF slags are being used in masonry mortars [30,[32], [33], [34]], in concrete mortar or as rapid hardening material [[35], [36], [37], [38], [39]], as mineral aggregate in preparing plaster [40], in paving soil-cement mix for roads [33,41], as liner material in landfill cover [[42], [43], [44], [45]], in embankment construction [46], in production of cement [24,27], as admixtures in ceramic building products or cement composites [47,48], reusing by injecting within the furnace [49,50], recycling itself in several steelmaking stages [51], in preparing geopolymers [[52], [53], [54]], as binder supplement in briquettes [55], as supplementary binder or cementing material [18,21,29,34,[56], [57], [58]], in hydraulic concrete having no or slight load bearing capacity [37], in preparing precast concrete [37], in producing hot mix asphalt or asphalt roadway recycling [20,59], as filler material in self-compacting concrete and in bituminous mixtures [[60], [61], [62], [63]], inclusion as a construction material in rigid or flexible composites [64], in replacement of limestone as neutralizing agent in bioleaching processes without any adverse effect on the bioleaching efficacy [65], as a treating material (mixed with treated residue of spent catalyst) to treat another industrial waste for stabilizing mercury [66], for removing phosphorus during side stream treatment in asphalt blend as filler and fine aggregates [26], in remediation of acid soil to check the effect in growth ability of plants in soil [39], in alkali activation process [35,[67], [68], [69], [70]], for carbon uptake and sand substitute [71], as adsorbent for removing detrimental elements from solutions, wear resistance [72], application in pavement [25].

1.2. Environmental issues related to LRF slag

In the era of modernization, newer types of waste bring more problems in managing or handling them. If mass LRF slag is directly dumped without concern, it may cause surface and groundwater contamination, health hazards like breathing problems, and so on. As a result, effective recycling or reuse of LRF slag has become an environmental issue as open stockpiling or direct disposal of LRF arises [38,58,73]. Though Guzzon et al. [74] and Branca et al. [31] stated that the physicochemical properties of LRF may cause dust emissions or particulate depositions in the atmosphere while handling for reuse. Najm et al. [38] ascertained LRF as a nonhazardous waste having no environmentally significant detrimental elements. Researchers [18,20,75,76] investigated the leachability of LRF and concluded that LRF possesses no environmental threats regarding leachability, which is encouraging in applying supplementary cementitious material or as aggregates in roadwork applications. The investigators are inspired to reuse these byproducts in various fields to reduce disposal-related problems [29,33,35].

1.3. Soil improvement with LRF slag

The application of steel slag or slag-based mixture for stabilizing soft clay or improving soil properties is a well-known practice in geotechnical engineering [[77], [78], [79], [80], [81], [82], [83], [84], [85], [86], [87]]. This practice also ensures proper management of the waste byproducts by reusing them rather than landfilling, reducing extra attention to LRF management. Previous studies indicate that scholars have emphasized using LRF slag to improve the properties of soft clayey soil. Chaiyaput & Ayawanna [88] examined the effect of LRF slag on the mechanical properties of clayey-sandy soil and weak lateritic soil. The authors observed the increase in CBR value with the increase in LRF slag proportion, while the swelling index was found to be decreased caused by a decrease in plasticity index. The LRF slag was used in stabilizing clayey soil, which improved the unconfined compressive strength of soil and the dynamic modulus with the addition of 10% and 15% of LRF slag [2]. Ortega-López et al. [89] worked on the long-term stimulated expansion of several ladle furnace slags and their utilization in soil stabilization, and the authors concluded that LRF slag can be used as a typical soil stabilizer, efficiently leading to the improvement of soil properties. Adding LRF slag in clayey soil increased the bearing capacity and compressive strength and reduced the plasticity index and free swelling [25,90]. Ortega-López et al. [89] and Manso et al. [90] examined the improvement of soft clay soil using LRF slag. Serjun et al. [75] aimed to evaluate the potentialities of LRF slag by reviewing its possible usage and describing various methods of recycling LRF slag as a construction material. The authors performed SEM/EDS investigation to check the hydraulic properties of LRF slag and concluded that this slag could be used as building and civil engineering material.

1.4. Sustainability and economic prospects of using LRF slag

Since the world is having difficulties due to shortages in energy resources, changes in global climate, and increase in material usage and waste generation, geotechnical engineers and researchers should think about sustainable energy and materials through recycling to solve these global issues [91]. Recycling LRF slag for improving soil for different geotechnical applications is an economical and environment-friendly approach as utilizing LRF ensures sustainable solid waste management along with providing alternative materials sources [92,93]. Song et al. [94] summarizes the use of steel slags as a sustainable construction material based on chemical and mineral compositions and the associated physical-chemical properties. According to Das et al. [95] steel slag can also be safely used with clinker and gypsum to produce cement since it is comparable to clinker and has a low iron concentration and a high calcium oxide content. Costa et al. [96] indicated that using steel slag instead of natural aggregates makes it possible to produce less polluted and less costly materials while conforming to the circular economy's standards, which helps promote sustainable solutions for the construction industry. The circular economy concept could be accomplished by incorporating the treatments and steering of slags' qualities into the overall process optimization system and reusing [97]. Hence, researchers suggest to sustainably replace aggregates by LRF as civil engineering materials due to their physical, chemical, and mineralogical characteristics [91,95,[98], [99], [100]].

To add to the existing research outcomes, this study attempted to use fine-grained ladle refined furnace (LRF) slag to replace commonly used stabilizers (cement, lime, etc.) to stabilize the clayey soil by performing various laboratory tests to determine the physical, mechanical, and strength properties. In this research, silt-sized LRF slag was used to make a better mixing with silt-sized soil particle, and hence there will be minimal pores in the LRF-soil mixture and better stabilization, while most researchers used coarser-sized LRF aggregates in soil. Researchers worked with lower percentages of LRF slags before, this research reported detailed test results with various percentages up to 35% of slag to check the optimum LRF content to stabilize the soil, ensuring the maximum utilization and recycling of LRF in soil improvement. Here, LRF slag was used independently as a cementitious material without any other additive to improve the settlement behavior of the soil mixed with LRF slag powder. This study has also regional importance as foundation soil stabilization independently using LRF in the study area, Sylhet, Bangladesh, was not done before, main novelty of this work.

2. Materials and methods

2.1. Materials

Clay soil was collected from a construction site in Sylhet City, where excavation works for foundation construction were ongoing. The soil samples were extracted by an excavator with the help of site labor from a depth just below the isolated footing level (about 6 ft below the top surface). Soil sample was sealed in polythene sacks to avoid moisture loss during carrying. As the humidity in study area is very high (≈90%) during sample collection and drying, the collected soil was oven-dried at 100 °C for at least 24 h and stored in the laboratory for further testing.

LRF slag was collected from a bulky steel manufacturing industry located in Chattogram, Bangladesh, named BSRM. LRF sample was oven-dried at 100 °C for at least 24 h and stored in air-tight conditions in the laboratory.

2.2. Methods

2.2.1. Sample preparation

The original oven-dried soil sample completely passed through No. 4 sieve, with 16.4% passing through No. 200 sieve (0.075 mm). When pulverizing with a wooden hammer with minimal effort, without breaking soil grains, 79.4% soil samples passed through No. 200 sieve (shown in Fig. 1(a)). The soil samples passing through No. 230 sieve (0.063 mm), finer than silt-sized (MIT soil classification), was only used for further analysis in this research to exclude unusually big particles like foreign materials. On the other hand, 100% of the collected dry LRF slag passed through No. 4 sieve, with 12.3% passing No. 200 sieve. When the LRF samples are dried at oven at 100 °C for 24 h and pulverized using a wooden hammer, without breaking the particle, 73.8% of LRF slag passed through No. 200 sieve (Fig. 1(b)). LRF slag was mixed with the clay soil starting from 5% to 35% with an increment of 5% for each sample on a dry-weight basis to make a clear assessment between the original soil and LRF-soil mixture. For every specimen test, the soil-LRF slag mixture was thoroughly mixed by taking the specified percentage of the LRF slag, and remolded samples were prepared for the intended tests following ASTM STP 599 [101].

Fig. 1.

Oven dried-pulverized samples: a) soil sample, b) LRF slag.

2.2.2. Atterberg limit tests

In geotechnical engineering, the Atterberg limit test plays an important role in the assessment of cohesive soil properties, i.e., how soil behavior changes with the changes in water content. Atterberg limit test was conducted to determine the liquid limit (LL), plastic limit (PL), and plasticity index (PI) of the soil sample and soil-LRF slag mixtures following ASTM D4318 [102].

2.2.3. Compaction test

Standard Proctor test was employed to determine the moisture content-dry density behavior of the soil and the soil-LRF slag mixtures according to ASTM D698 [103]. The test was performed using 600 kN-m/m³ effort in a 6-inch mold, and the sample was trampled into three layers with 56 blows in each layer every time. The compacted materials were weighed and regained to calculate the actual water content. The optimum water content corresponding to each sample maximum dry density (MDD) was then assessed from the moisture-density curves.

2.2.4. Unconfined compressive strength (UCS) test

In order to estimate the efficiency of LRF slag in soil stabilization, it is imperative to quantify the capability of LRF slag on the sample's strength gain in contrast to the strength of original soil sample. The unconfined compressive strength of the remolded soil samples was determined according to ASTM D2166 [104]. Firstly, the samples were compacted following ASTM D698 [103] at standard effort. Then the cylindrical soil sample core was extracted from the mold with a sampling tube and sample extruder. The prepared UCS specimens with a diameter of 38 mm and height of 76 mm were permeated in plastic bag for 48 h, and then the testing was carried out, and unconfined compressive strength of the soil samples was determined.

2.2.5. Consolidation test

With an intention to observe the compressibility behavior of the soil sample and soil-LRF slag mixtures, consolidation tests were performed according to ASTM D2435M [105]. From this experiment, the coefficient of consolidation (c_v), compression index (C_c), and coefficient of permeability (k) for all the samples were determined to examine the effect of the LRF slag in the clay soil. The incremental load was applied to the samples in 6 steps, creating stresses of 6.11 kPa, 12.22 kPa, 24.45 kPa, 48.9 kPa, 97.08 kPa, and 195.6 kPa, where 24-h interval were allowed to complete the primary consolidation of the specimen in each loading step.

2.2.6. California Bearing Ratio (CBR) test

The prepared soil samples of the original soil and soil-LRF slag mixtures were tested for the California Bearing Ratio (CBR) test following ASTM D1883 [106]. The CBR mold was 178 mm in height and 155 mm in diameter, while the loading plunger had a diameter of 50 mm. All the samples were compacted at their respective optimum moisture content and maximum dry density found in the compaction tests. The compacted soil samples were soaked in water for 48 h prior to loading. The load corresponding to penetration was recorded and plotted to determine the CBR value of all the samples. The maximum CBR value between 2.54 mm and 5.08 mm penetration of the plunger into the soil is taken as the final CBR value of the soil sample.

3. Results and discussion

3.1. Characterization of the materials

The bulk density and natural water content of the soil sample has been recorded as 1.96 g/cm³ and 18.1%, respectively. The basic physical properties of soil sample are presented in Table 1. The specific gravity of the soil is 2.67, while the specific gravity of LRF is 3.23. The physical properties of LRF slag are presented in Table 2. LRF was found as a whitish grey with a dry density of 1.5–2.5 gm/cm³ and a pH of 10–11, which is partially soluble in water. The compositions from analytical results of the clay soil and LRF slag are presented in Table 3 and Table 4, respectively. In the analytical test results of soil sample, the main compositions were aluminum oxide (Al₂O₃), silicon dioxide (SiO₂), ferric oxide (Fe₂O₃), potassium oxide (K₂O), carbon dioxide (CO₂), and a portion of titanium dioxide (TiO₂). LRF slag primarily consists of silicon dioxide (SiO_2), aluminum oxide (Al₂O₃), ferrous oxide (FeO), calcium oxide (CaO), and magnesium oxide (MgO), and manganese (II) oxide (MnO), where calcium oxide was the predominant element in the chemical compositions, and the rest was mostly silicon dioxide and aluminum oxide (Table 4).

Table 1.

Basic properties of soil.

Bulk density (gm/cm³)	1.96
Moisture content (%)	18.1
Specific gravity	2.67
Optimum moisture content (%)	16.5
Maximum dry density (gm/cm³)	1.77
Unit weight (kN/m³)	16.48
Soil type	clayey

Table 2.

Physical parameters of the LRF slag sample.

Appearance	Solid, granular powder
Color	Whitish grey
Odor	NA
Density	1.5–2.5 gm/ml
pH	10-11 in 1% solution at 25 °C
Melting point	Above 1500 °C
Explosive limit	Non-combustible
Solubility in water	Partially soluble
Inflammability	Non-inflammable

Table 3.

Compositions of soil sample (analytical results).

Test Parameters	Units	Results
Al2O3	%	27.27
SiO2	%	37.16
Fe2O3	%	10.23
K2O	%	5.12
TiO2	%	4.31
CO2	%	13.58
Pb	(mg/kg)	BDL
Cd	(mg/kg)	BDL
C	(mg/kg)	6.2
Dissolved salts	(mg/kg)	0.73
BDL- Below detection limit

Table 4.

Compositions of LRF slag sample (analytical results).

Test Parameters	Units	Results
SiO2	%	24.27
Al2O3	%	6.38
FeO	%	1.23
MnO	%	1.12
CaO	%	53.16
MgO	%	3.58
Pb	(mg/kg)	BDL
Cd	(mg/kg)	BDL
Cr	(mg/kg)	67.6
Ni	(mg/kg)	11.5
BDL- below detection limit

3.2. Atterberg limit

The changes in the consistency properties, i.e., Atterberg limits of clay soil after being mixed with LRF slag in different percentages, are presented in Table 5. As the LRF slag has a significant proportion of calcium oxide and silicates (Table 4) with very low plasticity, it helps avoid difficulties due to the expansive nature of clayey soils in water [61,107,108]. The liquid limit (LL), plastic limit (PL), and plasticity index (PI) of the original soil sample have been found as 41.32%, 21.8%, and 19.52%, respectively. According to the Casagrande Plasticity Chart shown in Fig. 2., the soil sample is classified as a high plasticity clay, positioned just above the low plasticity clay range. The result indicates from other tests that primarily LL and PI of the soil-LRF slag mixture specimens increased to some extent with the addition of 5% LRF slag. After that, LL began to decrease with an accumulation in LRF slag percentage, while the changes in PL were very low to reasonably constant. The maximum LL recorded for soil with 5% LRF slag was 52.17%, with corresponding plasticity index of 30.34%, followed by a decrease in both LL and PI with the increase in LRF percentages in the soil. The soil sample with 35% LRF slag shows the lowest LL of 36.53%, PL of 23.06%, and PI of 13.47%. Manso et al. [90] also experienced a similar reduction trend in the plasticity index of the soft clay soil stabilized with furnace slag. The reduction in LL and PI of the stabilized soil is due to the changes in bonding behavior of soil in presence of cementitious materials [89], which also changes the soil classification of the soil-LRF mixture into low plasticity clay, as shown in the plasticity chart in Fig. 2 [79].

Table 5.

Atterberg limit test results.

Soil sample	LRF slag (%)	Liquid Limit (%)	Plastic Limit (%)	Plasticity Index (%)
Original soil	0	51.32	21.80	29.52
Soil + 5% LRF	5	52.17	21.83	30.34
Soil + 10% LRF	10	47.23	21.92	25.31
Soil + 15%LRF	15	44.81	22.01	22.8
Soil + 20%LRF	20	42.58	22.06	20.52
Soil + 25%LRF	25	40.25	22.56	17.69
Soil + 30%LRF	30	38.85	22.82	16.03
Soil + 35%LRF	35	36.53	23.06	13.47

Fig. 2.

Atterberg limits presented in Casagrande Plasticity Chart for all the samples tested.

3.3. Compaction test

The dry density vs. water content graph obtained from the Standard Proctor test of the original soil sample and the LRF slag mixed samples are presented in Fig. 3. The OMC was found as 16.5% for the original soil, for which dry density becomes 1.77 g/cm³. As the LRF is added to the soil, the OMC to achieve the maximum dry density (MDD) rises. The addition of LRF slag in the clay soil sample increased the OMC substantially, which is due to high water absorption capacity of the LRF slag [2]. A rise in MDD with the addition of LRF slag percentages up to 20% has been observed, while for further addition, a decrease in MDD value has been recorded though the value was still higher than that of the original soil. So, in terms of dry density gain, the optimum LRF slag percentage was 20%, at which the dry density became 1.94 g/cm³. Osinubi [109] also observed an increase in MDD with the increment of slag percentages up to a certain amount, while the OMC increased. The higher water absorption characteristics of LRF also resulted in a rise in the Atterberg limit, leading the LRF-soil mixture to increase OMC, while the MDD of soil increased up to 20% LRF in the soil and then decreased. The reduced maximum dry density of the treated soil due to the further addition of LRF slag depicts the higher resistance to the resistive effort generated by the flocculated soil structure [79].

Fig. 3.

Dry density vs. water content for the soil samples.

3.4. Unconfined compressive strength (UCS)

The results of UCS tests for all the samples are presented in Fig. 4. The UCS value of the soil sample was noted as 742.76 kPa with a displacement of 4.79 mm at peak load. A significant increase in the UCS was observed with the addition of LRF slag in the soil. The UCS value becomes 1220.06 kPa, 1314.85 kPa, 1536.81 kPa, 2368.51 kPa, 1452.24 kPa, 1443.41 kPa, and 1334.11 kPa for the samples containing LRF slag of 5%, 10%, 15%, 20%, 25%, 30%, and 35%, respectively. There were also variations in the displacement of the soil sample at their peak load; the displacement becomes 5.15 mm, 4.53 mm, 5.78 mm, 6.15 mm, 7.62 mm, 6.11 mm, and 4.37 mm for the sample containing 5%, 10%, 15%, 20%, 25%, 30%, and 35% LRF slag, respectively. For the soil sample with 20% LRF, a 219% increment in UCS was achieved, which was 2368.51 kPa. After that, a reduction in UCS has been observed with further addition of LRF slag. The soil sample mixed with LRF slag showed resistance against compressive load up to an optimal content of LRF slag [85,87]. Therefore, the optimum LRF slag percentage was 20% from the UCS test results. In the formation of stabilized soils with adequate bearing capacity, the interaction of the reactive LRF materials with the clayey soils is undoubtedly beneficial and involves the use of a significant amount of LRF, which increases the demand for LRF slag recycling [26,87,89].

Fig. 4.

Unconfined compressive strength (UCS) comparison for all the samples.

3.5. California Bearing Ratio (CBR)

Fig. 5 and Table 6 describe the load vs. penetration behavior for all the samples under CBR tests. The soil samples containing LRF in different percentages showed resistance in load bearing, and enabling them to take a higher load for the same penetration than the original soil sample. For the original soil sample, the equivalent load for 2.54 mm penetration was found as 5.1 kN and 8.05 kN for 5.08 mm penetration of the plunger into the soil after CBR correction (Fig. 5). The original soil sample gives a CBR value of 37%, and for the soil sample with 5% LRF slag, the CBR value slightly increased to 39.45%. Further, an increase in LRF slag content up to 20% in the soil shows an increase in CBR value of the samples; with 20%LRF, CBR value became the highest, i.e., 54.40% (Table 6). Then a decrease in CBR value was observed by further addition of LRF slag in the soil, and this is because more percentage of slag decreased the dry density of the soil-LRF mixture, which is shown in Table 7. Aldeeky and Hattamleh [78] reported that 20% fine steel slag with plastic subgrade clay provided a significant improvement in CBR value. As the LRF shows cementitious behavior, it hardens the samples, the soil becomes more resistant to load after soaking, and the plunger penetration into the soil sample becomes less than that of the original soil sample, resulting in a greater CBR value [80,86]. According to the CBR test results shown in Table 7, 20% LRF is the optimum LRF slag content in the soil sample, as this content shows the greatest CBR value among all the samples tested [78].

Fig. 5.

Load vs. penetration of the different samples in CBR tests.

Table 6.

Load vs. penetration data of CBR test for all the samples tested.

Penetration (mm)	Load (kN)
	Original soil	Soil + 5% LRF	Soil + 10% LRF	Soil + 15% LRF	Soil + 20% LRF	Soil + 25% LRF	Soil + 30% LRF	Soil + 35% LRF
0	0	0	0	0	0	0	0	0
0.45	0.13	0.16	0.19	0.22	0.37	0.37	0.37	0.29
0.64	0.2	0.23	0.26	0.29	0.52	0.43	0.39	0.35
0.92	0.45	0.47	0.49	0.53	0.83	0.65	0.62	0.58
1.27	1.03	1.08	1.12	1.18	1.98	1.39	1.23	1.28
1.5	1.6	1.79	1.95	1.99	3.12	2.42	2.23	2.18
1.75	2.3	2.45	2.65	2.85	3.94	3.18	3.03	2.99
1.94	2.6	2.82	2.82	3.1	4.53	3.65	3.45	3.33
2.2	3.2	3.5	3.65	3.97	4.99	4.54	4.21	4.17
2.54	3.7	3.99	4.2	4.52	5.94	5.13	4.98	4.83
3.18	4.87	5.19	5.39	5.79	6.99	6.42	6.21	6.09
3.81	5.67	5.97	6.32	6.72	8.12	7.27	7.01	6.94
4.2	6.15	6.65	6.95	7.25	8.81	7.96	7.72	7.66
4.45	6.57	6.93	7.23	7.73	9.18	8.48	8.12	8.08
5.08	7.42	7.92	8.22	8.64	10.36	9.35	8.96	8.82
5.72	8.13	8.43	8.83	9.19	10.61	9.89	9.52	9.49
6.35	8.67	8.98	9.28	9.64	10.99	10.4	10.1	9.94
6.99	8.92	9.29	9.59	9.99	11.01	10.72	10.33	10.2
7.62	9.2	9.43	9.73	10.2	11.03	10.87	10.49	10.37
8.26	9.35	9.65	9.85	10.35	11.05	10.74	10.63	10.54

Table 7.

Observations of variations in soil properties with LRF.

LRF content (%)	0	5	10	15	20	25	30	35
OMC (%)	16.5	17.3	18.5	19.8	20.1	21.2	22.3	23.6
Dry density (g/cm3)	1.77	1.79	1.814	1.94	1.92	1.91	1.88	1.86
UCS (kPa)	742.76	1220.06	1314.85	1314.85	2368.51	1452.24	1443.41	1334.11
CBR value (%)	37.00	39.45	40.27	43.14	54.40	46.70	42.81	41.50

3.6. Consolidation parameters of samples in evidence of strengthening soil

Fig. 6 shows the void ratio, e vs. vertical stress, σ for different samples in all the loading stages used in the consolidation test. The slope of the steepest part of the e vs. σ graph (shown in Fig. 6) indicates the compression index, Cc, which is 0.358 for the original soil sample (Table 8). Compression index of the samples decreased with the increase in LRF percentages in soil up to 20% LRF and became the lowest value of 0.269. Compression index increased with the further addition of the LRF slag in the soil. Settlement of the soil samples due to incremental loading in the consolidation test is also shown in Table 8. The highest settlement was found to be 3.93 mm for the original soil sample. As the LRF content increased in the soil sample, the samples experienced lower settlement than the original soil. The soil sample showed a decrease in settlement with the increase in LRF content up to 20% in the soil and became the lowest at 2.73 mm. After that, further increase in LRF content results in a slight increase in settlement value. An increment in LRF content in the soil sample up to a certain level makes the soil samples less compressible and stabilized at the end of loading; hence, the settlement of the LRF-contained soil is less than that of the original soil. Rezaei-Hosseinabadi et al. [110] and Eyo et al. [111] both observed a lower compression index and settlement in the clay soil stabilized by cementitious materials, including steel slags, and alkaline metal based additives. The cementitious behavior of the LRF content in soil samples enhanced the stabilization, strength, soil-LRF interaction, and less void ratio, which resulted in improved soil samples having higher shear strength with less settlement [17,111]. The data shown in Table 8 reveals that 20% LRF is the optimum content for the soil samples used in this research.

Fig. 6.

Void ratio vs. Stress for all the tests carried out on samples.

Table 8.

Settlement and compression index for the soil samples from consolidation test.

Soil sample	Settlement (mm)	Compression index, Cc
Original soil	3.93	0.358
Soil with 5%LRF	3.72	0.327
Soil with 10%LRF	3.45	0.313
Soil with 15%LRF	3.11	0.290
Soil with 20%LRF	2.73	0.269
Soil with 25%LRF	2.95	0.283
Soil with 30%LRF	2.97	0.295
Soil with 35%LRF	2.99	0.298

3.6.1. Scanning electron microscopy (SEM) tests

The Scanning Electron Microscopy (SEM) analysis was executed by ESEM QUANTA200-FEI apparatus furnished with energy dispersion spectroscopy for chemical analysis (X-EDS Oxford INCA‐ 350) on the original soil, LRF slag, and 20% LRF slag mixed soil in order to understand the physical activities which take place at the sample surface. In the case of the original soil Fig. 7(a), irregularity among the particles has been noticed with many micropores, i.e., the particles are placed far apart with pore space among them. Again, the particles of the original soil sample are less transparent, although the surface of the particles seems to be smooth. For the LRF slag sample in Fig. 7(b), it is seen that the particles exhibit shiny, denser, and partly porous conditions. The surface of the LRF slag particles is much clearer than that of the original soil; the particles seem to be rough with less opacity. When the soil sample and LRF slag were mixed, the image (Fig. 7(c)) shows that the particles accumulated to make flocculation and formed bonds reducing the micropores, indicating the stabilization of the soil. Due to the presence of silicon, calcium, and aluminum in LRF slag, particles of the soil-LRF mixed sample showed interconnections and became denser [[112], [113], [114]]. As such, structural heterogeneity was noticed in the mixture of clay and LRF slag from the SEM image due to the interaction between the particles of clay and LRF slag [86].

Fig. 7.

Scanning electron microscopy (SEM) images: a) the original soil, b) LRF slag, c) Soil with 20% LRF slag.

The results of the energy dispersion spectroscopy (EDS) test for the original soil sample, LRF slag, and soil-LRF slag mixture are shown in Fig. 8(a) and (b) and 8(c), respectively, while the quantitative EDS data has been presented in Table 9. The EDS quantitative data shows that O, Si, and C are the predominant compounds for the original soil sample, while Ca, O, and B are the essential compounds in the LRF slag. Again, the soil sample with 20% LRF slag primarily contains O and Si, which indicates the interaction between soil particles and LRF slag [85,87].

Fig. 8.

EDS images: a) original soil sample, b) LRF, c) Soil+20% LRF.

Table 9.

SEM-EDS corresponding data from quantitative EDS results.

Original soil sample
Element Line	Net counts	Int. Cps/nA	Z	Weight %	Norm. Wt.%	Atom %	Formula
C K	2398	0.869	0.92	8.34	8.34	12.75	C
O K	30905	11.197	0.968	58.19	58.19	66.81	O
Al K	22498	8.151	1.076	10.06	10.06	6.85	Al
Si K	38448	13.93	1.05	17.3	17.3	11.32	Si
K K	3871	1.403	1.113	1.68	1.68	0.79	K
Ti K	739	0.268	1.196	0.43	0.43	0.17	Ti
Fe K	4776	1.73	1.205	4	4	1.31	Fe
Total				100	100	100
Ladle Refined Furnace (LRF)
B K	266	0.096	0.909	4.74	4.74	11.6	B
O K	1061	0.384	0.919	23.81	23.81	39.33	O
Mg K	211	0.076	0.988	0.44	0.44	0.48	Mg
Al K	699	0.253	1.023	1.05	1.05	1.03	Al
Si K	4007	1.452	0.997	5.07	5.07	4.77	Si
Ca K	49767	18.032	1.034	64.9	64.9	42.8	Ca
Total				100	100	100
Soil + 20% LRF
O K	33152	12.012	0.967	61.52	61.52	73.91	O
Si K	92435	33.491	1.049	37.71	37.71	25.81	Si
Ca K	282	0.102	1.089	0.14	0.14	0.07	Ca
Fe K	672	0.243	1.204	0.62	0.62	0.21	Fe
Total				100	100	100

4. Conclusions and recommendation

Reusing industrial waste byproducts to enhance the characteristics of soft soils is an excellent example of promoting environmental sustainability, which the current global demand for innovation desires. This study examined the capability of LRF slag in stabilizing the clay soil for potential use as foundation soil for low-rise buildings. However, before applying this technique in the field, economic analysis of soil stabilization using LRF has to be done in detail. According to the findings of this study, with various percentages of LRF slag mixed, clay soil shows significant improvement in different geotechnical properties. From the findings of this research, future investigators and other specialists will be benefitted by getting information concerning the efficiency of LRF slag in improving the properties of poor soil. The following conclusions can be drawn into attention.

• ⁃The Atterberg limit tests show that with the increase in LRF slag in the soil, both the liquid limit and plasticity index decreased. The soil samples transformed from original high plastic clay to low plastic clay with 35% LRF slag in the soil. A maximum liquid limit of 52.17% was recorded for the sample of soil-5% LRF and the corresponding plasticity index was 30.34%.
• ⁃In the compaction test, the optimum moisture content increases with the addition of LRF slag, while maximum dry density rises with the addition of 20% LRF slag in the soil mass.
• ⁃The unconfined compressive strength of soil increased by 219% due to the addition of 20% LRF in the soil and slightly decrease for the LRF content beyond 20% but always more than the original soil sample.
• ⁃California Bearing Ratio test showed that LRF slag-containing soil exhibited more resistance in load bearing and grabbed more load for equal penetration than the original soil sample. The soil sample containing 20% LRF increased the CBR value by 47% than that of the original soil.
• ⁃The consolidation test reveals that up to 20% LRF in the soil mass, compression index of soil decreased along with the decrease in consolidation settlement by 30%, which may solve consolidation-induced foundation settlement issue in clay soil.

Funding statement

This research was funded by SUST Research Center (Project ID: AS/2021/1/18), Shahjalal University of Science and Technology (SUST), Sylhet.

Data availability statement

Data will be made available on request.

Ethics approval and consent to participate

Not applicable.

Additional information

No additional information is available for this paper.

CRediT authorship contribution statement

Shriful Islam: Conceptualization, Formal analysis, Funding acquisition, Methodology, Project administration, Resources, Supervision, Writing – original draft, Writing – review & editing. Sabrin Ara: Funding acquisition, Methodology, Project administration, Resources, Supervision, Writing – review & editing. Junaidul Islam: Formal analysis, Investigation, Visualization, Writing – original draft.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:

Shriful Islam reports financial support was provided by Shahjalal University of Science and Technology. Sabrin Ara reports financial support was provided by Shahjalal University of Science and Technology. Shriful Islam reports a relationship with Shahjalal University of Science and Technology that includes: employment. Sabrin Ara reports a relationship with Shahjalal University of Science and Technology that includes: employment.